- 21 Jul, 2015 14 commits
-
-
Jamie Madill authored
In some cases, binding nullptr could query for the internal format of the null object, causing us to dereference null and crash. BUG=angleproject:1080 TEST=dEQP-GLES3.functional.multisample.fbo_* Change-Id: I592ab31dc1efadfac003c39051632fc2a0fff0cc Reviewed-on: https://chromium-review.googlesource.com/286853Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Tested-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
Passing a 2D texture type to a 3D texture function could result in an explosion. BUG=angleproject:1079 TEST=dEQP-GLES3.functional.negative_api.texture.compressedteximage3d Change-Id: Ibfcc4bdb334270f8fa27ebed6bcc3911986e7c7c Reviewed-on: https://chromium-review.googlesource.com/286852Reviewed-by:
-
Jamie Madill authored
We had a bug where we would require "stride * primCount" bytes when drawing from a buffer, when the last element would not require the full stride. The correct expression: "stride * (primCount-1) + size. BUG=angleproject:1086 TEST=dEQP-GLES3.functional.vertex_arrays.single_attribute.strides.* Change-Id: I7e2897c9d18b0e9849289d033f0ae2573237e4e5 Reviewed-on: https://chromium-review.googlesource.com/287320Tested-by:
-
Geoff Lang authored
Also make sure that EGL_EGLEXT_PROTOTYPES is always defined so that extension functions can be used without loading them. BUG=angleproject:970 Change-Id: I33fa3e8ed68aeda55ad69e1bc7826646f5e3ce29 Reviewed-on: https://chromium-review.googlesource.com/287161Reviewed-by:
-
Corentin Wallez authored
With this commit angle programs can be run under apitrace: when we use shared libraries, the apitrace driver shim shadows ANGLE and redirects the GL calls to the driver, which is not what we want. BUG=angleproject:892 Change-Id: I32535ba8db6202b0b5b285e86abd472366ee3d03 Reviewed-on: https://chromium-review.googlesource.com/287270Reviewed-by:
-
Corentin Wallez authored
Otherwise apitrace doesn't know the dimension of the framebuffer and cannot display it. BUG=angleproject:892 Change-Id: Ib8ad846ca80e4783ab77c96430bb5220df1a39ab Reviewed-on: https://chromium-review.googlesource.com/286881Reviewed-by:
-
Geoff Lang authored
Causing MSAN failures on Linux. This reverts commit 1915652e. Change-Id: Ib23bec16eab22288930be0b41186e54cd8d1f921 Reviewed-on: https://chromium-review.googlesource.com/287127Reviewed-by:
-
Corentin Wallez authored
This replaces the dlsyming of glX entry point with linking to libGL which in trun allows the ANGLE to be traced using apitrace. Re-land with a trivial compilation fix (MEMBER vs MEMEBER) BUG=angleproject:892 Change-Id: Ie89b6fe26eb80d3d8d48f49645d7a6254230bf5c Reviewed-on: https://chromium-review.googlesource.com/287143Reviewed-by:
-
Jamie Madill authored
This will clear the way to put the GLES3 tests on the bots. BUG=angleproject:1071 BUG=510822 Change-Id: I65fc917425531f6c080bcc862802d25af10cf265 Reviewed-on: https://chromium-review.googlesource.com/287164Tested-by:
Kenneth Russell <kbr@chromium.org>
-
Corentin Wallez authored
This reverts commit cddca5f204d9aa55c9810cf30c74997ca19a7f86. Change-Id: I07fa96a8b84308c77fb642a64f39fcacc471d96e Reviewed-on: https://chromium-review.googlesource.com/287123Reviewed-by:
-
Corentin Wallez authored
This replaces the dlsyming of glX entry point with linking to libGL which in trun allows the ANGLE to be traced using apitrace. BUG=angleproject:892 Change-Id: Ic065582401e0e1e3fce1329cde216220a1e086a8 Reviewed-on: https://chromium-review.googlesource.com/286880Reviewed-by:
-
Geoff Lang authored
These were causing assertions in several WebGL2 conformance tests. BUG=angleproject:882 BUG=angleproject:1084 BUG=483282 Change-Id: I7651c359a528c95469dcbfb5da2715ae4616b49b Reviewed-on: https://chromium-review.googlesource.com/286592Reviewed-by:
Zhenyao Mo <zmo@chromium.org> Tested-by:
-
Geoff Lang authored
This is an attempt to fix use-of-uninitialized-value errors. Example failure: https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_msan_rel_ng/builds/68/steps/cc_unittests%20%28with%20patch%29/logs/EnlargedTextureWithAlphaThresholdFilter.GL BUG=angleproject:1044 Change-Id: I5906c67a55da553e2e5fd15320a5dd4186644e87 Reviewed-on: https://chromium-review.googlesource.com/287191Reviewed-by:
-
Kimmo Kinnunen authored
Make all fragment shader out variables require location layout qualifier. Previously, the last variable did not need a location layout qualifier if the previous variables had those. TEST=angle_unittests BUG=angleproject:1070 Change-Id: Ifb66ee52b811409f5278eaad330d6cd9b8ea059f Reviewed-on: https://chromium-review.googlesource.com/287100Tested-by:
Kimmo Kinnunen <kkinnunen@nvidia.com> Reviewed-by:
-
- 20 Jul, 2015 19 commits
-
-
Geoff Lang authored
Despite many extensions being written against specific versions of the GL spec, some drivers (Intel) expose the extensions in much earlier versions. This leads to issues where the extension string would be present but the entry points were not loaded. BUG=angleproject:1038 Change-Id: Ia87bb3f038f310f6bc3e711a0b2d82717b67ae70 Reviewed-on: https://chromium-review.googlesource.com/286525Reviewed-by:
-
Geoff Lang authored
It should be equivalent to GL_MAX_VARYING_COMPONENTS for ES2. BUG=angleproject:1038 Change-Id: Ic2fe6c27a0ebd6259d70f8d28edb7d4bdd5d0c35 Reviewed-on: https://chromium-review.googlesource.com/286526Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:883 Change-Id: I3cdf88391e77a26a77e120de0fd32296a2b079d1 Reviewed-on: https://chromium-review.googlesource.com/286822Reviewed-by:
-
Jamie Madill authored
BUG=angleproject:762 Change-Id: Ic0f862ac901eb793266e68ce26cc9be67d5cb12d Reviewed-on: https://chromium-review.googlesource.com/286862Tested-by:
-
Jamie Madill authored
The ES3 spec is clear that these calls should no-op, see page 187, "If a buffer is not present, then a Clear directed at that buffer has no effect." BUG=angleproject:1078 TEST=dEQP-GLES3.functional.fbo.msaa.2_samples.* Change-Id: Ic521f55785d10065e2ffe2e98a68c9d7da0c2c71 Reviewed-on: https://chromium-review.googlesource.com/286776Reviewed-by:
-
Jamie Madill authored
Causes a warning on Windows: compiler\translator\ValidateOutputs.cpp(37): warning C4804: '>' : unsafe use of type 'bool' in operation BUG=angleproject:1070 This reverts commit 140941d0. Change-Id: Ieed42cdda22f17c6e15c38ee1c059184869e6919 Reviewed-on: https://chromium-review.googlesource.com/286820Reviewed-by:
-
Kimmo Kinnunen authored
Make all fragment shader out variables require location layout qualifier. Previously, the last variable did not need a location layout qualifier if the previous variables had those. TEST=angle_unittests BUG=angleproject:1070 Change-Id: I3763b8ca38b1e14ee8456a54592c01e0fd89692c Reviewed-on: https://chromium-review.googlesource.com/286101Tested-by:
-
Jamie Madill authored
This will allow us to eventually make a presubmit script for ANGLE, and also will help new contributors who aren't as familiar with the style guide. One outstanding issue is the lack of proper breaking before array initializers. We don't use arrays all over the place, but this will likely require a patch for clang. BUG=angleproject:762 Change-Id: I4702b1fcdfa10b7682b07316be59ec67bcf84943 Reviewed-on: https://chromium-review.googlesource.com/285901Reviewed-by:
-
Jamie Madill authored
There was a typo in the vertex shader cache which could trigger an overflow bug. TEST=dEQP-GLES3.functional.ubo.random.scalar_types.* BUG=angleproject:1077 Change-Id: I5652cf9675155b627f84531e09c01b42e29278fc Reviewed-on: https://chromium-review.googlesource.com/286775Reviewed-by:
-
Jamie Madill authored
The blit refactoring mislabeled one shader as GL_RED instead of GL_RED_INTEGER. BUG=angleproject:1076 TEST= dEQP-GLES3.functional.fbo.render.shared_depth_stencil.* Change-Id: I262b218c72d955a88a2d2c00b470d6a4f3568451 Reviewed-on: https://chromium-review.googlesource.com/286774Reviewed-by:
-
Jamie Madill authored
These were causing validation errors in the dEQP FBO tests. See Table 3.2 in the ES3 spec. BUG=angleproject:1074 TEST=dEQP-GLES3.functional.fbo.completeness.renderable.texture.color0.* Change-Id: Ifbf9199c54060eb3fc1a1dac7bbd74e19e013c61 Reviewed-on: https://chromium-review.googlesource.com/286773Reviewed-by:
-
Jamie Madill authored
This was causing several dEQP tests with index data to fail. BUG=angleproject:1073 TEST=dEQP-GLES3.functional.buffer.map.read_write.render_as_index_array.* Change-Id: Ic231a561202fbcda499b7584905b80252a7407de Reviewed-on: https://chromium-review.googlesource.com/286772Reviewed-by:
-
Jamie Madill authored
We were neglecting to transpose the types here, so were getting incorrect registers counts for rows and columns. BUG=angleproject:1072 TEST=dEQP-GLES3.functional.shaders.matrix.* Change-Id: Ia0ab89bbba8b9e2d8f39e84ef4e8ad3c03bb118d Reviewed-on: https://chromium-review.googlesource.com/286771Reviewed-by:
-
Jamie Madill authored
Using the SRV format could lead to using R32F for D32F formats, which is incorrect. BUG=angleproject:1075 TEST=dEQP-GLES3.functional.fbo.blit.* Change-Id: Ic81d51b888897a7f7c79e2ba1a24ab69f18d981c Reviewed-on: https://chromium-review.googlesource.com/286770Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:1044 Change-Id: I7c4bde80448b2cd64921b809a3a0dfb37c4e0f47 Reviewed-on: https://chromium-review.googlesource.com/286643Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:916 Change-Id: I5be7460b737095eedb886be3e8978765c0940757 Reviewed-on: https://chromium-review.googlesource.com/286784Reviewed-by:
-
Corentin Wallez authored
Using that idiom on an empty vector results in an undefined behavior that I have seen cause a vector to get length -1 with clang 3.6 on Linux. BUG= Change-Id: I3bb8d8884efa29a17672b458263067644dbf0fec Reviewed-on: https://chromium-review.googlesource.com/286740Reviewed-by:
-
Geoff Lang authored
Cascades to remove serials from many objects. BUG=angleproject:970 Change-Id: I0a74a14519da3203cd5df7ae0fa19f83393951ce Reviewed-on: https://chromium-review.googlesource.com/286554Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:970 Change-Id: I33120ec8121dc1265360fbbe61103e03e593af3f Reviewed-on: https://chromium-review.googlesource.com/286550Reviewed-by:
-
- 17 Jul, 2015 7 commits
-
-
Geoff Lang authored
BUG=angleproject:970 Change-Id: I63a7d8363f90018827c5d8482c9d98f3fa61c573 Reviewed-on: https://chromium-review.googlesource.com/286553Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:970 Change-Id: I1606df42e6423fe84d3ab69029a37224adb89a9b Reviewed-on: https://chromium-review.googlesource.com/286552Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:970 Change-Id: I4fae29766ee76512375872535dbaf2d31a379de0 Reviewed-on: https://chromium-review.googlesource.com/286551Reviewed-by:
-
Geoff Lang authored
BUG=angleproject:1044 Change-Id: I0e7bee366398159ae6b87a820f36cb51c05cb064 Reviewed-on: https://chromium-review.googlesource.com/280362Reviewed-by:
-
Jamie Madill authored
Change-Id: I24ea67f354191697d8a577c8b21ccc1689dd3536 Reviewed-on: https://chromium-review.googlesource.com/286016Tested-by:
-
Corentin Wallez authored
This will help address a bug in glReadPixels performed after a resize. BUG=angleproject:1050 Change-Id: Ie204dec229582ec0e990c6f7fa22cfd44048d46e Reviewed-on: https://chromium-review.googlesource.com/286250Reviewed-by:
-
Corentin Wallez authored
This reverts ac2582d3 which was a speculative improvement. It turns out that resizing before the swap is important as it lets the driver know that it should resize the default framebuffer, making the following frame have a default framebuffer of the right size. BUG=angleproject:1050 BUG=angleproject:892 Change-Id: I84ec3728226fde9fb208f91c02a3d5a1ab71aef8 Reviewed-on: https://chromium-review.googlesource.com/286252Reviewed-by:
-
